Robust Query Processing in Co-Processor-accelerated Databases

Robust Query Processing in Co-Processor-accelerated Databases
复制标题

协处理器加速数据库中的稳健查询处理

DOI:
10.1145/2882903.2882936
复制
发表时间:
2016
期刊:
Proceedings of the 2016 International Conference on Management of Data
影响因子:
--
通讯作者:
Jens Teubner
Jens Teubner
中科院分区:
--
文献类型:
--
作者:
Sebastian Breß;Henning Funke;Jens Teubner

文献摘要

参考文献

被引文献

相似文献

技术限制使得异构计算设备的使用远远超过了学术好奇心。事实上,使用此类设备被广泛认为是实现用户迫切需要和期望的应用程序加速的唯一有前途的方法。然而,建立一个强大的,高效的查询引擎异构协处理器环境仍然是一个重大的challenge.In本文中,我们确定了两个影响,限制性能的情况下协处理器资源变得稀缺。当查询的工作集不适合协处理器的数据缓存时,会发生缓存颠簸,导致性能下降高达24倍。堆争用发生时,多个操作并行运行在一个协处理器上,当他们积累的内存占用超过主内存容量的协处理器,查询执行速度减慢了6倍。数据驱动的操作符放置避免了可能有害的数据移动;查询截断限制了协处理器内存的使用,从而避免了争用。这种组合方法(数据驱动的查询切碎)在协处理器上实现了鲁棒且可扩展的性能。我们验证了我们的建议与我们的开源GPU加速数据库引擎CoGaDB和流行的星星模式和TPC-H基准。
Technology limitations are making the use of heterogeneous computing devices much more than an academic curiosity. In fact, the use of such devices is widely acknowledged to be the only promising way to achieve application-speedups that users urgently need and expect. However, building a robust and efficient query engine for heterogeneous co-processor environments is still a significant challenge.In this paper, we identify two effects that limit performance in case co-processor resources become scarce. Cache thrashing occurs when the working set of queries does not fit into the co-processor's data cache, resulting in performance degradations up to a factor of 24. Heap contention occurs when multiple operators run in parallel on a co-processor and when their accumulated memory footprint exceeds the main memory capacity of the co-processor, slowing down query execution by up to a factor of six.We propose solutions for both effects. Data-driven operator placement avoids data movements when they might be harmful; query chopping limits co-processor memory usage and thus avoids contention. The combined approach-data-driven query chopping-achieves robust and scalable performance on co-processors. We validate our proposal with our open-source GPU-accelerated database engine CoGaDB and the popular star schema and TPC-H benchmarks.
DOI: 10.12694/scpe.v11i4.663
发表时间: 2010
期刊: Scalable Comput. Pract. Exp.
影响因子: --
作者:
Jie Cheng
通讯作者: Jie Cheng
DOI: 10.1145/2588555.2610507
发表时间: 2014-06
期刊: Proceedings of the 2014 ACM SIGMOD International Conference on Management of Data
影响因子: --
作者:
Viktor Leis;P. Boncz;A. Kemper;Thomas Neumann
通讯作者: Viktor Leis;P. Boncz;A. Kemper;Thomas Neumann
细胞宽带引擎上的矢量化数据处理
DOI: --
发表时间: 2007
期刊: International Workshop on Data Management on New Hardware
影响因子: --
作者:
S. Héman;N. Nes;M. Zukowski;P. Boncz
通讯作者: P. Boncz
DOI: --
发表时间: 2014
期刊: SIGMOD Conference
影响因子: --
作者:
Tomas Karnagel;M. Hille;M. Ludwig;Dirk Habich;Wolfgang Lehner;Max Heimel;V. Markl
通讯作者: V. Markl
Red Fox:GPU 上关系查询处理的执行环境
DOI: --
发表时间: 2014
期刊: IEEE/ACM International Symposium on Code Generation and Optimization
影响因子: --
作者:
Haicheng Wu;G. Diamos;T. Sheard;M. Aref;S. Baxter;M. Garland;S. Yalamanchili
通讯作者: S. Yalamanchili